Fairness Resource Allocation in Blind Wireless Multimedia Communications

Liang Zhou,Min Chen,Yi Qian,Hsiao-Hwa Chen
DOI: https://doi.org/10.1109/TMM.2013.2237895
IF: 7.3
2013-01-01
IEEE Transactions on Multimedia
Abstract:Traditional -fairness resource allocation in wireless multimedia communications assumes that the quality of experience (QoE) model (or utility function) of each user is available to the base station (BS), which may not be valid in many practical cases. In this paper, we consider a blind scenario where the BS has no knowledge of the underlying QoE model. Generally, this consideration raises two fundamental questions. Is it possible to set the fairness parameter in a precisely mathematical specific -fairness resource allocation schememanner? If so, is it possible to implement a specific -fairness resource allocation scheme online? In this work, we will give positive answers to both questions. First, we characterize the tradeoff between the performance and fairness by providing an upper bound of the performance loss resulting from employing -fairness scheme. Then, we decompose the -fairness problem into two subproblems that describe the behaviors of the users and BS and design a bidding game for the reconciliation between the two subproblems. We demonstrate that, although all users behave selfishly, the equilibrium point of the game can realize the -fairness efficiently, and the convergence time is reasonably short. Furthermore, we present numerical simulation results that confirm the validity of the analytical results.
What problem does this paper attempt to address?